Millions of people are exposed to drinking water contaminated with arsenic, and extensive epidemiological evidence has demonstrated that these exposures can cause cancer. In fact, at high concentrations the levels of risk exceed those of any other known environmental carcinogen. At relatively low exposures such as those commonly found in the U.S., cancer risks from lifetime exposure could be above 1 in 1000 people, even at the newly proposed drinking water standard of 10 micrograms/L. In certain susceptible subpopulations, such as those who smoke or have poor diets, the risks may be even greater. Furthermore, our studies in Chile suggest particularly high risks for those drinking arsenic-contaminated water as children. Unfortunately, cancer risks at low exposures are uncertain since risk estimates to date involve extrapolation from high dose levels to lower exposures where the shape of the dose-response relationship is unknown. Such extrapolations are highly controversial. We therefore propose a population-based case-control study to assess the association of lung cancer with low to moderate levels of arsenic in drinking water. Current evidence indicates that lung cancer may be responsible for more deaths due to ingested arsenic than all other cancer sites combined, including bladder cancer. It is therefore particularly important to obtain a clear picture of the dose-response relationship for this cancer. The study area includes Kings County, California, and six counties in Nevada. These counties incorporate the largest population in the U.S. exposed to water supplies containing between 50 and 100 micrograms/L of arsenic. Most other water supplies in the study region contain less than 5 micrograms/L and thus provide a marked contrast in exposure. A total of 271 lung cancer cases diagnosed in the study area between 2002 and 2004 will be identified with rapid case ascertainment from local hospitals. Tumor biopsies will be archived for a potential subsequent study of DNA alterations. Random digit dialing and the rolls of the Health Care Financing Administration will be used to identify two controls for each case, frequency-matched by age and sex. Telephone interviews of all study subjects will be conducted to gather information on lifetime residential history and drinking water sources which will be used in conjunction with water arsenic measurements to construct exposure histories. A strength of the study is that exposure can be reliably ascertained retrospectively since it is largely dependent on residential history. Information on cigarette smoking will also be obtained and synergistic effects with arsenic assessed. Dietary information, medical history, and demographic data will be collected and analyzed for potential susceptibility factors. The proposed study has over 83 percent statistical power to detect a relative risk of 1.7, the risk predicted by linear extrapolation from high dose studies. The study has public health importance since finding the hypothesized relative risk would identify important health effects from low levels of exposure, whereas not finding increased risks would contribute to assurance about public health protection from the new drinking water standard.